Journal of Chromatography A, 1123 (2006) 92–97

Determination of chlorine dioxide in water by

gas chromatography–mass spectrometry
Ho-Sang Shin a,b,∗ , Dong-Gyun Jung c
aDepartment of Environmental Education, Kongju National University, Kongju, Republic of Korea
b Abuse Drug Research Center, Kongju National University, Kongju, Republic of Korea
c Department of Environmental Science, Kongju National University, Kongju 314-701, Republic of Korea

Received 15 January 2006; received in revised form 22 April 2006; accepted 28 April 2006
Available online 22 May 2006

Abstract
A sensitive gas chromatographic method has been established for the determination of chlorine dioxide in water. With weak basic conditions (pH
9.0), chlorine dioxide reacts with iodide to form iodine, which reacts with 2,6-dialkylphenol to form 4-iodo-2,6-dialkylphenol. The volatile organic
derivative was extracted with ethyl acetate, and then measured by gas chromatography–mass spectrometry (GC–MS). The reaction of the active
proton of 2,6-dialkylphenols (2,6-dimethylphenol, 2,6-di-isopropylphenol and 2,6-di-tert-butylphenol) with iodine was tested, and compared to
each other in terms of reactivity and stability of the derivatives. 2,6-dimethylphenol showed rapid reaction with iodine, and its derivative was stable
for 2 weeks. The detection limit of chlorine dioxide in water was about 1.0 ng/mL, and the calibration curve showed good linearity with r2 = 0.998.
The existent concentration of chlorine dioxide in water was calculated from multiplying the concentration calculated from the calibration curve of
4-iodo-2,6-dimethylphenol by 0.544. The method was sensitive, reproducible and simple enough to permit the reliable analysis of chlorine dioxide
at the low ng/mL level in water.
© 2006 Elsevier B.V. All rights reserved.

Keywords: GC–MS; Chlorine dioxide; Determination; Water

1. Introduction by-products (chlorite and chlorate ions) [1]. Typically, chlorine

Chlorine dioxide was discovered in 1811 by Sir Humphery
Davy, who called the green–yellow gas euchlorine [1]. Chlo-
rine dioxide (ClO2 ) is one of the readily available alternatives
to chlorine for disinfection and oxidation during the treatment
of drinking water [1–3]. Chlorine dioxide readily dissolves in
water to form a solution that is biocidal to a wide range of
microorganism. It is also known that chlorine dioxide undergoes
disproportionations in water, producing chlorite and chlorate.
The major advantage of chlorine dioxide is that the disinfec-
tion process does not generate large quantities of halogenated
by-products that are common with chlorination [4,5]. Another
advantage is that chlorine dioxide significantly reduces odor and
color [1]. The major disadvantage of chlorine dioxide, however,
has been the potential health hazards of the resulting inorganic
∗ Corresponding author. Tel.: +82 416 50 8811; fax: +82 416 50 8811.
E-mail address: hshin@kongju.ac.kr (H.-S. Shin).
dioxide is used at a concentration level between 0.1 and 2.0 mg/L
and it may exist in the concentration of low ng/mL after the
treatment. Water treated with chlorine dioxide contains not only
chlorine dioxide, but also chlorite and chlorate. A monitoring
method of chlorine dioxide is required a sensitivity enough to
detect to low ng/mL without the interference of chlorite and
chlorate.
Many methods based on different principles have been pro-
posed for the determination of chlorine dioxide. Spectropho-
tometry has been widely used for the determination of chlorine
dioxide in drinking water by using many kinds of dyes [6–16].
Iodometry [17], DPD method [18], ion-chromatography [19–21]
and HPLC [22] have also been reported as techniques for the
determination of chlorine dioxide, but their sensitivities are not
enough to detect to low ng/mL usually. In order to enhance the
response of the compounds, gas diffusion flow-injection analysis
(FIA) [23–25] have been performed.
In our previous paper, it was described that phenol derivative
was formed by the displacement reaction of iodine with active

0021-9673/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.chroma.2006.04.089
 H.-S. Shin, D.-G. Jung / J. Chromatogr. A 1123 (2006) 92–97
proton at para-position of 2,6-dimethylphenol, and the resulting
derivative was analyzed by GC–MS [26]. Our aim in this study
is to develop an analytical method by using the reactivity of
chlorine dioxide with iodide and iodine with 2,6-dialkyl phenols.
2. Experimental
2.1. Reagents
Potassium iodide and iodine were obtained from Sigma
(ST. LOUIS, MD, USA). 2,4,6-Trichlorophenol (TCP), 2,6-
dimethylphenol (2,6-DMP), 2,6-diisopropylphenol (2,6-DIPP),
2,6-di-tert-butylphenol (2,6-DTBP) and 4-iodo-2,6-dimethyl
phenol (4-iodo-2,6-DMP) were obtained from Aldrich (Milwau-
kee, WI, USA). Ethyl ether, ethyl acetate and sodium sulfate
were used as reagents and solvents.
Chlorine dioxide gas was produced by the reaction between
sodium chlorite (Wako, Japan) and hydrosulforic acid, and then
trapped in ultra pure water at 0 ◦ C and stored in a cool dark
place at 2 ◦ C, followed by quantitation of this chlorine dioxide
solution using the iodomeric titration [16] just before analysis
and diluted to the various concentrations with ultra-pure water.
The pure water used in this study was purified by a Milli-Q-
Reagent-Grade water system (ZD20) and had a resistivity of
over 17 M
. pH of 0.5 mM Tris solution was adjusted to 9.0
with 0.2 M HCl.
2.2. Derivatization and extraction
Ten milliliters of water containing 0.01–50 ␮g of chlorine
dioxide was placed into a 20 mL glass-stoppered test tube. Two
millilters of Tris–HCl buffer (pH 9.0) and 10 ␮L of potassium
iodide solution (1% in pure water) were added to the solution
and agitated for 1–2 min. Then a 10 ␮L of 2,6-DMP solution (5%
in MeOH) was added to the solution and shaken mechanically
for 10 min at room temperature. Hundred microliters of 2,4,6-
TCP (50 ␮g/mL in MeOH) as an internal standard was added
to the solution. The sample was extracted with 3 mL of ethyl
acetate by mechanical shaking for 10 min, and then about 10 mg
of ascorbic acid were added to the solution to eliminate residual
chlorine dioxide following agitating for 30 s. The two phases
were separated by centrifugation (5 min at 1500 × g) and the
organic phase was transferred into a 20 mL glass stoppered test
tube and evaporated to 100 ␮L of residual solvent with nitrogen
stream. The solution was transferred into a V-shape in auto vial,
and a 2 ␮L sample of the solution was injected in the GC system.
2.3. Calibration and quantitation
Calibration curve for the linearity test was established by
adding 0.02, 0.05, 0.1, 0.25, 0.5, 1.0, 2.5, 5.0, 10.0, 20.0 and
50.0 ␮g of 4-iodo-2,6-dimethyl phenol standard and 5.0 ␮g of
2,4,6-TCP (ISTD) in a 10 mL of distilled water. The solutions
were extracted with ethyl acetate after the redox reaction with
iodide and the subtitution reaction with 2,6-DMP, and the extract
was injected in the GC system. Calibration curve for the quan-
titation was prepared by adding 0.02, 0.05, 0.1, 0.25, 0.5, 1.0,
93
2.5, 5.0, 10.0, 20.0 and 50.0 ␮g of 4-iodo-2,6-DMP standard and
5.0 (g of 2,4,6-TCP in a 10 mL of distilled water. The solutions
were extracted with ethyl acetate, and the extract was injected in
the GC system. The ratio of the peak area of standard to that of
internal standard was used for the construction of the calibration
curve.
2.4. Gas chromatography–mass spectrometry (GC–MS)
All mass spectra were obtained with an Agilent 6890/5973N
instrument. The ion source was operated in the electron ion-
ization mode (EI; 70 eV, 150 ◦ C). Full-scan mass spectra (m/z
40–450) were recorded for analyte identification. An HP cross-
linked PEG capillary column (InnoWax) was used. Samples
were injected in the split mode with a split ratio of 1:10. The
flow rate of helium as carrier gas was 0.9 mL/min. The GC oper-
ating temperatures were: injector temperature, 240 ◦ C; transfer
line temperature, 240 ◦ C; oven temperature, programmed from
100 ◦ C (held for 3 min) at 10 ◦ C/min to 240 ◦ C (held for 3 min).
The ions selected for quantification by SIM were m/z 97 and 196
for TCP (internal standard) and m/z 121 and 248 for 4-iodo-2,6-
DMP.
3. Results and discussion
3.1. The chlorine dioxide–iodide reaction
For the GC analysis of chlorine dioxide, the derivatization
is necessary. The redox reaction between chlorine dioxide and
iodide ion has been interested from many researchers [27–30].
With weak basic conditions (pH 9.0), chlorine dioxide reacts
with iodide to form iodine (I). This reaction serves as a basis for
the iodometric determination of ClO2 .
2ClO2 + 2I−
I2 + 2ClO−
2 (I)
Every mole of chlorine dioxide in water was converted into
0.5 moles of iodine by the redox reaction.
3.2. The reactivity of iodine and 2,6-dialkylphenol
Phenol derivatives are formed by the displacement reac-
tion of active proton at para-position of 2,6-dialkylphenol as
a nucleophilic attack of the more electronegative iodine. 2,6-
Fig. 1. Time course of the reaction of 2,6-DMP, 2,6-D-i-PP and 2,6-D-t-BP with
iodine.
94 H.-S. Shin, D.-G. Jung / J. Chromatogr. A 1123 (2006) 92–97

Fig. 2. Chromatogram of the extract of (A) reagent water, (B) water spiked with chlorine dioxide (10 ng/mL) and water spiked with chlorine dioxide (20 ng/mL).
1 = 2,6-DMP; 2 = 2,4,6-TCP (500 ng/mL); 3 = 4-iodo-2,6-DMP.
 H.-S. Shin, D.-G. Jung / J. Chromatogr. A 1123 (2006) 92–97
Dialkylphenols (2,6-DMP, 2,6-DIPP, 2,6-DTBP) have an active
proton, which can be substituted with iodine.
The reaction of the active proton of 2,6-dialkylphenols (2,6-
DMP, 2,6-DIPP and 2,6-DTBP) with iodine was tested, and
compared to each other in terms of reactivity and stability of
the derivatives. The reaction rates of 2,6-DMP, 2,6-DIPP and
2,6-DTBP with iodine were determined by the detection of
the substituted products; 4-iodo-2,6-dialkylphenols at reaction
times of 1.0, 5.0, 10, 20, 40 and 60 min (Fig. 1). As results,
2,6-DMP and 2,6-DIPP showed very rapid reaction with iodine
(Fig. 1). Complete reaction takes place in about 1 min at room
temperature, provided that a sufficiently high concentration of
2,6-DMP is present in the reaction mixture. No significant varia-
tion in reaction yield was noted over this time period. Otherwise,
2,6-DTBP was converted to 4-iodo-2,6-di-t-butyl phenol, which
increased until 20 min.
To evaluate the stability of the derivatives, the experiment was
repeated by analyzing the extracts stored at 4 ◦ C for 2 weeks.
The mean percentual stability of 4-iodo-2,6-dialkylphenol after
2 weeks varied from 3.0% (for DMP) to 3.8% (for DIPP).
As a results, 4-iodo-2,6-DMP derivatized by 2,6-DMP
showed superior reactivity and stability compared to those by
the other methods.
The substitution reaction between iodine and 2,6-DMP was
used for the determination of chlorine dioxide in water. With
weak basic conditions (pH 9.0), chlorine dioxide reacts with
iodide to form iodine by reaction (I), which consequently reacts
with 2,6-dialkylphenol to form 4-iodo-2,6-dialkylphenol and
iodide ion by reaction (II).
I2 + 2, 6-dialkylphenol → 4-iodo-2, 6-dialkylphenol + I− (II)
Every mole of iodine formed from two moles chlorine dioxide
in water was converted into a mole of 4-iodo-DMP by substi-
tution reaction (II). Eventually, two moles of chlorine dioxide
in water were converted into a mole of 4-iodo-dialkylphenol by
redox and substitution reaction (I) and (II).
3.3. Extraction
Ethyl acetate was found to be efficient for the extraction of
4-iodo-2,6-DMP from aqueous solution. Ethyl ether generally
contains antioxidants of phenol form, which may react with
iodine to iodo-substituted phenols. The residual chlorine diox-
ide or iodine in water can be extracted with the solvent, and
break capillary column during the separation. Therefore, the oxi-
dants remaining in water after the redox reaction were reduced
with ascorbic acid before the extraction. The extract was directly
injected into the GC after concentration.
3.4. Mass spectrometry
The mass spectrum of 4-iodo-2,6-DMP shows a molecular
ion and base ion at m/z 248, and the diagnostic ions at m/z 121,
103 and 91. The ion at m/z 121 is from the cleavage of iodide
from molecular, and the ions at 91 and 103 were from the losses
of [2CH3 + I] or [H2 O+I] from the molecular ion.
95
3.5. Chromatography
For the GC separation of the derivative, the use of the polar
stationary phase was found to be efficient. The column was
stable over more than 1000 injections without notable change
of the separation characteristics. Chromatograms are shown in
Fig. 2. Separation of the derivatives and internal standard from
the background compounds of water was very good. There were
no extraneous peaks observed in a chromatogram of blank water
at the retention times of 15.34 and 17.94 min.
3.6. Linearity and calculation
A calibration curve was obtained from the extraction after the
addition of 4-iodo-2,6-DMP and ISTD in water. The regression
line of peak area ratios of 4-iodo-2,6-DMP to internal standard
on concentration using a least-squares fit demonstrated a linear
relationship with y = 0.9539x − 0.0159 and r2 = 0.9977, where
x is 4-iodo-2,6-DMP concentration (␮g/mL) and y is the peak
area ratio of the analyse to internal standard. Another calibration
curve was obtained from the redox reaction and the extrac-
tion after the addition of chlorine dioxide and ISTD in water.
The line of best fit of chlorine dioxide is y = 1.7685x + 0.0162
(r2 = 0.9984), where x is chlorine dioxide concentration (␮g/mL)
and y is the peak area ratio of 4-iodo-2,6-DMP to internal stan-
dard.
From two calibration curves, the ratio of the slope of 4-iodo-
2,6-DMP (a1) to that of chlorine dioxide (a2) is 0.539, where
a1/a2 means the experimental conversion factor to evaluate the
concentration of chlorine dioxide in water (CClO2 , mg/L) with
the concentration calculated from the calibration curve of 4-
iodo-2,6-DMP (CI-DMP , mg/L).
Theoretically, two moles of chlorine dioxide in water were
converted into a mole of 4-iodo-DMP. The concentration of chlo-
rine dioxide in the sample was calculated according to the fol-
lowing formula. CClO2 = CI−DMP × 0.544, where 0.544 means
the theoretical conversion factor to evaluate the concentration
of chlorine dioxide. The theoretical conversion factor shows the
small variation in comparison to the experimental conversion
factor.
Consequently, the existent concentration of chlorine dioxide
in water is calculated from CI-DMP × 0.544.
3.7. Precision and accuracy
The reproducibility of the assay was very good, as shown in
Table 2. For four independent determinations at three concen-
Table 1
Within-run precision and accuracy
Concentration added (ng/mL) Found concentration
Results Mean ± SD (RSD)
50 59, 59, 56, 55 57 ± 2 (3.9)
100 94, 87, 101, 102 92 ± 4 (4.6)
500 479, 488, 476, 475 480 ± 6 (1.3)
SD: standard deviation; RSD: relative standard deviation.
96 H.-S. Shin, D.-G. Jung / J. Chromatogr. A 1123 (2006) 92–97

Table 2
Interference effect of ions and organic compounds
Matrix Added concentration Spiked Comp (mg/L) Found (ng/mL)
of ClO2 (ng/mL)
Results (%) Recovery (%)

100 – 98 ± 3 (3.1) 98
Pure water
500 – 486 ± 14 (2.9) 97
100 – 91 ± 5 (5.5) 91
500 – 481 ± 26 (5.4) 96
Surface
500 ClO2 − (0.5), ClO3 − (0.5) 484 ± 24 (5.0) 97
Watera
500 OCl− (0.5) 1,178 ± 57 (4.8) 236
500 OCl− (0.5), Tris (0.5 mM) 566 ± 28 (5.0) 113
a Surface water contains inorganic ions (Na+ 7.00 mg/L, K+ 2.78 mg/L, Mg2+ 4.48 mg/L, Ca2+ 4.81, Cl− 0.9 mg/L, NO − 1.2 mg/L, HCO − 50.0 mg/L and SO 2−
3 3 4
2.5 mg/L) and organic compounds (BOD 2.3 mg/L).

trations of 50, 100 and 500 ng/mL, the coefficient of variation

was less than 5% as shown in Table 1.

3.8. Sensitivity

The combination of low background, high derivatization

yield and the high abundance of molecular ion of the deriva-
tive permit their determination in water at concentrations well
below those reported previously. Method detection limit of chlo-
rine dioxide was 1.0 ng/mL based upon an assayed water volume
of 10 mL. Method detection limit was defined by a minimum
signal-to-noise ratio of 3 and coefficients of variation for repli-
cate determinations (n = 5) of 15% or less. Because only a small
portion of the sample was used for each determination, the sen-
sitivity could be greatly improved by increasing sample size.
3.9. Interference
In order to examine the interference effect of ions and organic
compound, the method developed was applied to determine
chlorine dioxide in natural surface water, which contains inor-
ganic ions and organic compounds. Also, the mixture of ClO2 ,
ClO2 − , ClO3 − and OCl− in natural water was analyzed to deter-
mine the interference effect of the chlorine species. The results
were given in Table 2. It is suggested that ions such as chlorite,
chlorate, chloride, carbonate, sulfate, nitrate and metal ions did
not interfere the determination of chlorine dioxide in water. But
hypochlorite ion made the redox reaction with iodide ion and
interfered the analysis.
Watanabe et al. [24] used a Tris reagent solution at pH 9.0 to
eliminate the interference of hypochlorite, based on chloramines
production from the reaction between Tris and hypochlorite.
Tris reagent solution in this study eliminated about 88% of the
interference of hypochlorite, but did not completely.
3.10. Application
The time-course change of chlorine dioxide in water under
or without sunshine was demonstrated as an application of the
developed method. The results are given in Fig. 3. In water,
chlorine dioxide decreased very deeply in the sunshine than in
the dark condition.
Fig. 3. Degradation pattern of chlorine dioxide in the sunshine and in the dark.
4. Conclusion
This proposed method determines very sensitively chlorine
dioxide in large excess of chlorite and chlorate ions. The detec-
tion limit of chlorine dioxide was 1.0 ng/mL in water. The
method developed here is simple, sensitive and suitable for the
determination of low concentration of chlorine dioxide in water.
References
[1] E.M. Aieta, P.V. Roberts, J. AWWA 78 (1986) 62.
[2] J. Huang, L. Wang, N. Ren, X.L. Liu, R.F. Sun, G. Yang, Water Res.
31 (1997) 455.
[3] C. Korn, R.C. Andrews, M.D. Escobar, Water Res. 36 (2002) 330.
[4] C.-Y. Chang, Y.-H. Hsieh, S.-S. Hsu, P.-Y. Hu, K.-H. Wang, J. Hazard.
Mater. B79 (2000) 89.
[5] C.-Y. Chang, Y.-H. Hsieh, Y.-M. Lin, P.-Y. Hu, C.C. Liu, K.-H. Wang,
Chemosphere 44 (2001) 1153.
[6] F. Mantisi, J.-P. Gautier, Eur. PAT. Appl. EP 985929, (2000) 9.
[7] C. Elleouet, F. Quentel, C.L. Madec, Rev. Sci. Eau 12 (1999) 561.
[8] G. Wang, H. Chen, L. Yuan, Anal. Lett. 34 (2001) 2485.
[9] C. Hue, W. Gaizhen, Y. Li, Anal. Lett. 30 (1997) 1415.
[10] D.L. Sweetin, E. Sullivan, G. Gordon, Talanta 43 (1996) 103.
[11] T. Watanabe, C.Q. Lin, Y. Yoshimura, K. Uchiyama, K. Ohsawa, K.
Imaeda, Anal. Sci. 10 (1994) 35.
[12] N. Imaizumi, Y. Nakahara, K. Suzuki, K. Oikawa, Chem. Lett. 8 (1993)
1333.
[13] T. Watanabe, C.Q. Lin, S.M. Chen, Y. Yoshimura, K. Uchiyama, K.
Ohsawa, K. Imaeda, Anal. Sci. 8 (1992) 207.
[14] K. Oikawa, Soda to Enso 42 (1991) 491.
[15] B. Nowack, U. Buchs, H.-P. Kaiser, U. von Gunten, Proc. AWWA Water
Qual. Technol. Conf. 61 (1999) 1145.
[16] Z.L. Jiang, B.M. Zhang, A.H. Liang, Talanta 66 (2005) 783.
 H.-S. Shin, D.-G. Jung / J. Chromatogr. A 1123 (2006) 92–97
[17] E.M. Aieta, R.V. Roberts, M. Hernandez, J. Am. Water Works Assoc.
76 (1984) 64.
[18] A.T. Palin, Water Sewage Works 107 (1960) 457.
[19] B.A. Jale, Proc. AWWA Water Qual. Technol. Conf. 15 (1987) 1055.
[20] J. Sullivan, M. Douek, J. Chromatogr. A 804 (1998) 113.
[21] T. Watanabe, T. Idehara, Y. Yoshimura, H. Nakazawa, J. Chromatogr. A
796 (1998) 397.
[22] H.G. Poovey, R.J. Rando, J. Liq. Chromatogr. 18 (1995) 161.
[23] H. Chen, Y. Fang, L. Yuan, T. An, Lab. Rob. Autom. 11 (1999)
157.
97
[24] T. Watanabe, T. Ishii, Y. Yoshimura, H. Nakazawa, Anal. Chim. Acta
341 (1997) 257.
[25] J. Gu, Y. Jian, L. Ji Feng, Microchim. Acta 148 (2004) 171.
[26] H.S. Shin, Y.S. O-Shin, J.H. Kim, J.K. Ryu, J. Chromatogr. A 732
(1996) 327.
[27] W.C. Bray, Z. Phys. Chem. 54 (1906) 575.
[28] R.E. Huie, P. Neta, J. Phys. Chem. 90 (1986) 1362.
[29] I. Fabian, G. Gordon, Inorg. Chem. 36 (1997) 2494.
[30] V. Csordas, B. Bubnis, I. Fabian, G. Gordon, Inorg. Chem. 40 (2001)
1833.